Will mobile phones and computing devices drive a continued ramp of CMOS into SoCs?
Ed.: This is the second of an occasional series by the authors of the 2017 iNEMI Roadmap. This information is excerpted from the Roadmap, which is available now from iNEMI (http://community.inemi.org/content.asp?contentid=51).
For the purposes of the iNEMI Roadmap, Portable & Wireless devices are defined as “high-volume consumer products for which cost is the primary driver, including handheld battery-powered products driven by size and weight reduction.” This sector contains such electronics devices as computers, tablets, mobile phones and smartphones, and portable gaming devices.
Within the portable product sector, there are fundamentally different product types. Wireless-based products such as mobile phones tend to have a greater percentage of discrete components to fine-tune the frequency matching and filtering functions. Processor-based products such as gaming devices use a greater proportion of ICs and fewer discrete components to achieve their functionality. This is reflected in the lead count per component. Design of processor-based products usually presents more challenges in the design of the interconnect substrate because of the routing between high I/O devices. However, the difference between these product types is narrowing due to convergence and electronic component integration. Although handheld products are becoming dominated by those with wireless connectivity, their circuit design has increasingly become more digital in nature, as a result of converging applications into the device and the increasing “digitization” and continuous integration of many previously analog design approaches.
Cost, power, bandwidth and form factor of these devices continue to be driving factors for acceleration of integration of silicon and system capabilities into single package or single die.
Mobile phone convergence. The convergence of mobile phones with computing and entertainment devices in effect enforces the further ramp of CMOS (complementary metal-oxide semiconductor) into SoCs (system-on-chip) both in homogeneous and heterogeneous technologies. The classical mixed technologies in 2D SIP (system-in-package) and MCP (multichip package) forms are now evolving and growing to more of a 3D MCP, with a new set of electrical, mechanical and thermal challenges. 3D stacking of chips will be mostly dominated by three key components: memory, application processor and communication silicon. They will leverage through-silicon via (TSV) and through-mold via in integrated, mixed technologies in single packages, as well as chip-on-chip (CoC) and chip-on-interposer over the next five years. Of course, the commercial challenges associated with 3D stacking, especially in the world of ASIC and memory integration, are expected to remain challenging.
Consignment, material, need for liquidity of memory size and even supply chain and test challenges will remain ramp barriers, as they all impact the bottom line. In the 2D space, under the “thin” initiative, bare die attach (BDA), Si interposer-based more complex “substrate” integration with embedded active and passive elements will come out of the experimental phase, and by 2017-19 a number of platform components will be mounted as bare die, mostly driven by thin form factor handheld and wearable devices, including phablets, cellphones, tablets and wearable consumer products.
On the other side of the spectrum, a number of technology leaders will continue pursuing the monolithic approach to integrate mixed technologies, especially for low- and medium-cost segments. The quest for higher power and performance CMOS PAs (power amplifiers) and the integration of analog and RF with traditional digital circuitry remains technically and commercially challenging. In the high-end segment, the discrete component will continue to dominate and meet the promise of quality and performance consumers are willing to pay for.
Displays. The desire for interoperability of devices and seamless computing will lead to a consumer electronics (CE) convergence of devices, with far fewer operating systems in play and a converged user interface working across devices with a set of applications (such as gaming, photo and video) that cross the boundaries of tablets, smartphones, TVs and set-top boxes.
Displays will migrate to 4K displays on the highest-end and QHD/full HD for the portable wireless markets, including smartphones/tablets and portable computers/2-in-1s. There will be more OLED-based displays at the highest-end smartphone (<7" screen size color displays).
All panels in smartphone/tablets and 2-in-1s will be touch-enabled, with advanced integration of touch sensors within the display gaining popularity.
The industry has a strong desire to go to OLED (organic light-emitting diode), but needs better emissive life, reliability and stability. The industry also has a strong desire for drop-resistant displays in the consumer handheld space. If production volume OLEDs can move rapidly enough away from their present rigid substrates toward plastic, they could provide much more design flexibility and eliminate risks associated with glass breakage. In addition, OLEDs are projected to provide faster response times, wider viewing angle, and decreased assembly thickness. Low-cost handsets will create a demand for disruptive display technologies such as electrophoretic displays (EPD), including flexible displays with electrophoretic front planes. Technologies such as Sharp’s announcement in 2012 of the IGZO (indium gallium zinc oxide) display will also contribute panels for devices sized between 7" and 32". Such technologies promise the benefits of increased battery life, as backlights would not require as great a brightness factor as previously.
The market demand for thinner devices with easier-to-view, larger graphics is creating an interest in “windowless” displays. In this configuration, the air gap between the window (glass or plastic typically attached to the case and above the display subassembly) is eliminated. Alternatively, the window itself can be eliminated and the display exposed. This will require improving the mechanical durability of the display subassembly. Designs like this already exist in digital still cameras, but consumer expectations of mobile phone durability will require a more ruggedized solution. As wearable devices begin to proliferate, smaller, thinner and lighter waterproof displays that are resistant to shock and drop events will see greater demand. Another issue is displays with good clarity and brightness indoors, as well as in direct sunlight. Any fix to these performance issues must not increase costs or power consumption. Nor should there be an increase in display thickness, and, ideally, no bezel. Form-factor minimization will dictate that the screen size be the size of the device, particularly on handheld and wearable devices.
Touch sensor applications. Incorporation of more features into devices is creating the need for additional input devices to use in interacting with the features. An example of this is touch-sensitive playback controls located within the display window of a mobile phone handset with MP3 playback capabilities.
Other applications for touch sensors are expected, with an increased need to provide intuitive feedback to substitute for the tactile “click” of a metal dome switch. Haptic feedback from the device to the user, as well as multi-touch and multi-swipe gesture panels, is ideal in consumer devices. Haptic feedback provides a user experience similar to a mechanical keyboard, and allows both simple and complex multi-touch experience – to the extent users feel as if they were holding parts of actual objects in CAD, gaming and other interactive environments. Micro-displays will emerge as devices continue to get smaller. These could be placed in a separate device such as a head-mounted display or incorporated directly into the device as a mini-projector. Projection-based devices will be needed, as camera sensors and perceptual computing will enable a bridge from the projector to true user interaction.
Work has begun on the 2019 Roadmap. iNEMI membership is not required to participate in the roadmap. Those interested may check the iNEMI website (http://community.inemi.org/2019_rm) for additional information, including a list of chapters and participation requirements.
is a principal engineer at Intel (intel.com) and chaired the Portable and Wireless Product Emulator Group (PEG) of the 2017 iNEMI Roadmap;